Estimate Residual Spillover From a Known-Negative Mask

Thin, in-memory wrapper over the batched pair estimator, for callers that already know which events are target-negative and do not need the estimator to infer it.

Usage

estimate.residual.spillover(
  unmixed,
  source,
  targets,
  negative.mask,
  threshold.source,
  threshold.target = NULL,
  spread.var = NULL,
  neg.var = NULL,
  ...
)

Arguments

unmixed

Numeric matrix, cells x fluorophores, already-unmixed abundances. Must contain a column named source and one column per entry of targets.

source

Character scalar, the name of the source fluorophore's column in unmixed.

targets

Character vector, the names of the target fluorophores' columns in unmixed to estimate residual spillover into.

negative.mask

Logical vector, length nrow(unmixed), TRUE for events already known to be target-negative. Used directly in place of the batched estimator's own negative-event inference.

threshold.source

Numeric scalar or vector of length nrow(unmixed), the source fluorophore's own per-event positivity boundary.

threshold.target

Optional numeric matrix, cells x fluorophores, containing at least the columns in targets, giving each target's own per-event positivity boundary. When NULL, every event is treated as below threshold before negative.mask is applied.

spread.var

Optional numeric vector, length length(targets), the source's contribution to each target's spillover-spread variance. Defaults to zero for every target when NULL.

neg.var

Optional numeric vector, length length(targets), each target's negative-population variance. When NULL, computed as stats::mad()^2 on each target's column of unmixed.

...

Additional arguments passed through to .fix.envelope.slope.batch().